Get to know Nerve Cloud: Big data gets smarter

Working with big data poses some common challenges. Nerve Cloud, a newly developed platform by QuantumBlack, solves them – giving clients transparency and control.

Three years ago, McKinsey acquired analytics firm QuantumBlack to grow the Firm's data and advanced analytics capabilities. One of QuantumBlack’s latest tools is Nerve Cloud – a reliable, scalable way of deploying cloud infrastructure, giving users transparency and control while working with off-the-shelf tools. The Nerve Cloud platform has supported more than 300 analytics studies.

Sam Bourton, QuantumBlack’s co-founder and Chief Technology Officer, talks here about what makes the Nerve Cloud platform distinctive, and how it has helped clients with complex issues.

* * *

Can you give us a refresher on what QuantumBlack does, and how it works with McKinsey clients?

At QuantumBlack we help companies identify and capture performance improvement opportunities by applying machine learning and AI to their data. The insights we derive help explain, optimize, and predict outcomes related to large, challenging, strategic problems using data and algorithms.

We take the client’s data, then create models to identify what the most informative drivers of success and failure are in their organizations. That could be anything from projects, products or assets, to factories, vehicles, mines, or oil rigs. We’re often looking at up to 60 siloed data sets that have never been joined.

For example, looking at several years of data from an engineering firm in the oil and gas industry, we can see what’s influenced the variance of outcomes in their projects. We can then solve for specific issues – such as what the optimal allocation of engineers to projects would be, maximizing for cost, quality, or time.

Can you tell us specifically what the Nerve Cloud platform does and how it works? Is there a reason why the word "Nerve" was chosen, or why it is particularly apt?

Nerve Cloud is part of a suite of niche QuantumBlack tech assets that we use to support and enhance our protocol for end-to-end analytics. This gives our clients access to a large service catalog of analytics tools, best-of-breed commercial off-the-shelf and open-source products – as well as our own tech assets.

The platform was designed and built to provide a single, secure portal to allow our teams to create an entire tech infrastructure environment and analytics workbench in the cloud in minutes. It gives our teams a fine-grained and transparent view of all of the services they can create.

The name ‘Nerve’ is apt because we think of an organization’s data landscape as its nervous system. We send tentacles out into that nervous system, interpret the signals from the noise, and generate intelligence from it.

There are a lot of companies doing data analytics. What makes the Nerve Cloud platform distinctive?

It’s very customizable, very quick, and very secure. Nerve Cloud is 100% infrastructure as code, which means that we can create instances of secure cloud environments incredibly quickly. And, at the same time, we can customize them for the specific needs and to reflect the specific investments in IT and infrastructure that our clients have made.

The platform allows us to create many hundreds of complete stealth-contained and isolated environments, which means that our clients’ data is never in the same area. There's no possibility of leaking the data between them.

The second distinctive element is that Nerve Cloud can build on a variety of cloud platforms. We currently have an Amazon version, but we've also worked on Azure, Google Cloud, on-premise, and private clouds.

Nerve Cloud integrates within a larger toolkit to provide a comprehensive ecosystem for analytics transformations. Besides Nerve Cloud, we also have frameworks for operationalizing AI into production environments, and a collaboration studio to allow teams of data engineers and data scientists to co-create data models and ontologies to support mulitiple analytics use cases.

What’s a particularly impactful engagement that you have worked on?

We've used Nerve Cloud on more than 300 engagements in the last 18 months. They're all impactful in different ways. [McKinsey alum] Vas Narasimhan, CEO of Novartis, has talked about how the company is using Nerve Cloud. The platform tracks every data point on more than 500 clinical drug-testing trials and uses analytics software to predict potential problems in the studies.

Another really exciting recent example comes from a heavily regulated industry. We couldn’t use Nerve Cloud hosted on the McKinsey AWS platform, because the client couldn’t allow the data to leave their organization. We were able to use a 100% infrastructure-as-code version to create an entire network topology on the client’s AWS cloud platform with virtual private clouds, private and public subnets, secure access roles and profiles, logging and auditing, instrumentation and monitoring, a secure firewall, and a complete security perimeter around it – from a single line of code, in 45 minutes.

This didn’t just give our team a secure environment to work on their data. It left behind a long-term, valuable tech asset and capability that the client can go on using themselves independently.

Can you talk about an instance where Nerve Cloud was able to solve a particularly problematic issue for a client?

We've used Nerve Cloud very effectively in client engagements where we need to get an environment up and running very, very quickly – faster than the client’s IT team might be able to do it when constrained by their corporate IT policies. And then we're able to transition all the data and code and logic to the client's own environment.

Nerve Cloud has also been useful for rapid prototyping and piloting. QuantumBlack manages the environment for a pilot phase, and then transitions all the code, data and analytics models back to their own environment for production.

Another example occurs when we're working with multiple clients, and we need to be a trusted broker of the data. We’ve used this in M&A engagements, where a group of companies that are merging needed to create a central secure data repository. McKinsey and QuantumBlack created the aggregated views and applied analytical models to them, providing only the aggregated views back to the clients and maintaining a robust security and compliance framework.